Ethics training programs framework: five design decisions and what to build first

A training program is a set of design decisions, and most of them are made by default. Who gets what, who delivers it, how often, what it is made of, and what happens to the questions it generates.

This is the design. It assumes you have an annual module that nobody remembers, which is the usual starting point.

What to take away

Decision one: what it is made of

Not principles. Cases from your own organization, anonymized, with the outcome removed so the group has to decide.

Collect them from three places: the questions managers were asked in the last two years, whatever the reporting channel received in category terms, and the incidents that reached the people function. Twenty cases is more than enough and most organizations can find them in a fortnight.

Each case is one paragraph and ends at the moment of decision. It has at least two defensible answers. An obvious answer teaches nothing and signals the session is theater.

Where a case turns on a genuine conflict of duties, a shared vocabulary helps. For most groups, a compact introduction to ethical decision making is enough of one.

Decision two: who gets what

Segment by exposure. Three or four groups is usually right.

The universal layer should be short, twenty minutes, and mostly about the route rather than about ethics. Everything else goes to the group that faces it. A finance analyst does not need the entertainment rules and will conclude, correctly, that the session was not built for them.

Decision three: who delivers it

Managers deliver the segment material to their own teams, with a prepared pack: three cases, the rules that apply, and the answers to the questions they will be asked.

This is the decision most organizations get wrong, because central delivery is easier to arrange and easier to report. But a session delivered by a stranger from head office teaches that conduct is a head office subject. A session delivered by the person who assigns your work and writes your review teaches that it is part of the work.

Managers will resist, and the honest answer is that the pack makes it a forty-minute job. Some will do it badly. A manager doing it badly still communicates more than a polished video does, and the ones who do it badly are themselves a finding.

The specialist function's job is to build the packs, run the intake loop, and handle the cases managers cannot answer.

Decision four: cadence

Annual everything is the default and it is the reason nothing sticks.

Short and frequent beats long and annual. Twenty minutes a quarter, on one case relevant to that team, run in an existing meeting. New joiners get the universal layer in their first fortnight, from their manager, not from a system.

Then two trigger-based sessions that matter more than the calendar ones. After any incident that the program should have prevented, run the case, anonymized, in the affected area. And before any period of concentrated pressure, such as a quarter end that everyone knows will be difficult, run the case about pressure.

Decision five: the intake loop

Every session generates questions. The loop is what turns them into next quarter's material and into changes in the rules.

The manager writes down every question asked and sends them, unedited, to the program owner. The owner reviews them monthly. A question asked three times in different teams is a rule that is unclear, and it goes to whoever owns the rule. A question nobody can answer is a gap and becomes a case.

Without this, the curriculum ages and the program becomes a repeated broadcast. With it, the material stays about the things people are actually uncertain about, which is the only content that holds attention.

What the program cannot do

It cannot make a rule apply to a top performer. It cannot make a reporting channel safe. It cannot survive an incentive that rewards the opposite, and it will be blamed for failures that belong to those three.

Say this explicitly to whoever sponsors the program, at the start, in writing. A training function that accepts responsibility for outcomes it does not control will spend its budget being blamed.

The honest scope is narrower: people know the route, know the rules that apply to their work, and have practiced a decision once before making it for real. Why the surrounding conditions matter more than the curriculum is set out in ethics training programs, and the incentive problem specifically in leadership ethics.

What to build first

If you are starting from nothing, in order: the manager pack for receiving a concern, the universal twenty minutes about the route, and the intake log. Those three take a month and cover most of the value.

An outside reviewer looking at a program after something has gone wrong asks whether training was tailored to the actual risks and whether it changed anything, rather than how many people completed it. That is the standard set out in the Department of Justice's guidance on evaluating corporate compliance programs.

Designing to that standard from the start costs no more than designing to a completion rate. The document the training makes usable is covered in codes of conduct, and what the person on the receiving end is actually weighing is in employee ethics.

Common questions

How long should the whole thing take per person per year?

Under two hours for most people, spread across four short sessions. Programs that need a day a year are usually delivering content, not practice, and the content is the part nobody retains.

What if our managers are the problem?

Then a central program will not fix it and will make it look handled. Deal with the manager layer first, which is a leadership question rather than a training one.

Do we need an external provider?

For building packs and facilitating difficult sessions, sometimes. For delivering the routine material, no, and outsourcing it undoes the main design decision above.